By Miyabi Endo

Yomiuri Shimbun Staff Writer

The manufacturing industry is turning its attention more and more to artificial intelligence as it aims to pass on skills and techniques to younger employees. Instead of getting the employees to "learn by watching," the industry, which faces an aging workforce and labor shortages, is employing trial and error to determine how to conduct training in a way that is practical and efficient.

Tablets

A work system based in the metaverse, a virtual space on the internet, was unveiled to the press at Hitachi Ltd.'s research and development center in Kokubunji, Tokyo. Speaking into a tablet, Takayuki Akiyama, a senior researcher at Hitachi, asked: "The sensor is indicating a high temperature. What should I do?" The AI responded: "There is likely not enough coolant. Please manually open the coolant valve." It then guided the user through the necessary steps.

The AI determines the optimal work procedure based on the equipment's technical drawings and trouble-shooting history. Specific operating methods for the equipment are displayed in the metaverse, making it possible not only for inexperienced workers but also industrial robots to use the system.

In verification tests conducted at Renesas Electronics Corp.'s factory, the work performance of unskilled workers improved by about 30%. "It is important to create a system that efficiently shares the skills of skilled workers," Akiyama said.

Utilizing virtual reality

Similar to Hitachi, Meidensha Corp., a major heavy electrical equipment manufacturer, is also attempting to efficiently transfer skills by "visualizing" them. As early as this autumn, the company will introduce a training system that uses virtual reality to recreate a manufacturing site with transformers and switchboards in a 3D space. "This will allow employees to learn anytime, anywhere," a company representative said. "We want to ensure that skills continue to be passed on amid concerns about labor shortages."

Shift from OJT

This new training approach offers potential for increased efficiency compared to on-the-job training (OJT), which has long been the mainstay method of sharing skills between workers. NEC Corp.'s subsidiary NEC Facilities Ltd. opened a space last year aimed at training facility managers for factories. The space features equipment identical to that found in actual factories, such as the clean rooms required for semiconductor manufacturing. Generative AI analyzes videos of work in progress and automatically produces daily reports and other documents pointing out areas for improvement and other matters. Traditionally, it took about 12 years to become a skilled worker, but that time can be reduced to six years, the company said.

According to the Economy, Trade and Industry Ministry's White Paper on Monozukuri (manufacturing), 10.46 million people were employed in the manufacturing industry in 2024, a decrease of about 10% over the past 20 years. Companies in the industry that said they have problems developing human resources and skills came to 85.3%, with more than 60% of them citing "a shortage of personnel to provide guidance."

----

This article is from The Yomiuri Shimbun. Neither Dow Jones Newswires, MarketWatch, Barron's nor The Wall Street Journal were involved in the creation of this content.

YDN-m0000130412-1

(END) Dow Jones Newswires

July 22, 2025 06:08 ET (10:08 GMT)

Copyright (c) 2025 The Yomiuri Shimbun